Anthropic accuses Chinese AI firms DeepSeek, Moonshot AI and MiniMax of using unauthorized distillation of its Claude models through 24,000 fraudulent accounts, generating over 16 million exchanges: The US AI lab's allegations expose a widely debated training technique, sparking controversy over acceptable boundaries in model development. https://www.scmp.com/tech/tech-war/article/3344499/anthropics-distilling-charges-against-chinese-firms-expose-ai-training-grey-area #China #TechNews #ChineseTech #AI #DeepSeek #Moonshot